Regulation of gene expression in humans is precisely controlled in order to ensure proper response to environmental and developmental cues. Indeed, aberrant gene expression leads to myriad problems, most notably cancer. While much of this regulation is imposed by activators binding specific DNA sequences, they typically require additional "co-activator" complexes to communicate their signals to the transcriptional machinery. We propose to analyze the structural (using cryo-EM) and functional characteristics of the CRSP (also called Mediator) co-activator complex in order to understand how it regulates transcription. Specifically, we will examine how CRSP-dependent activation of gene expression can be repressed by the "ARC-L specific" (cdk8, cyclin C, Med230, and Med240) polypeptides. We will also characterize the tight association of the CRSP-RNA polymerase II binary complex. This analysis will establish specific CRSP subunits and domains that are essential for its function and thus identify potential targets for therapeutics. The packaging of eukaryotic genes into chromatin provides an additional, finely-tunable mechanism of transcriptional regulation. In particular, wrapping of DNA into nucleosomes can result in repression of gene expression by making either promoter or enhancer sequences inaccessible to proteins required to achieve the necessary levels of transcription. ATP-dependent chromatin remodeling complexes alter nucleosomal structure and therefore the accessibility of nucleosomal DNA, without removal or modification of histones. Here we propose to characterize the architecture of four related human and Drosophila remodeling complexes belonging to the SWI-SNF family by Cryo-EM. Our aim is to identify common features defining an enzymatic core, as well as to pinpoint evolutionary, functional, and specificity-related differences between the complexes. To gain information on the mechanism of remodeling, we will study the interaction of these complexes with recombinant nucleosomes and in the presence of different nucleotides. The final goal is to be able to generate a mechanistic model of how these ATP-driven molecular machines affect nucleosome structure.